OA3-AM24-ST-19 - Diversity of Causes of C Discrepancies Identified by Automated Testing

Blood specimens from first-time African American blood donors that were tested on both serologic and molecular high-throughput platforms were evaluated for C antigen typing discrepancies between the platforms. The results of investigation are presented.

Study

Design/Methods:

During a 24-month period (Feb. 2022 – Feb. 2024), 36,722 samples were tested on Ortho Vision (OV) analyzer (Quidel Ortho) for C, c, E and K antigens, then used for DNA extraction and low-resolution RBC genotyping using HemoID DQS (DQS, Agena Bioscience) which predicts 45 red cell antigens. Results from both platforms are automatically inputted into the donor database. Discrepancies between test results generated an automated alert and were investigated by tube testing and medium-resolution RH genotyping using RHCE and RHD BeadChip (Werfen).

Results/Findings:

Of the 36,722 donors tested, there were 26 (0.07%) C antigen discrepancies.  In 22 cases, tube testing was performed; all were concordant with gel testing on the OV. Thirteen (50%) were due to either weak antigens (RHCE*CE) or variant antigens with anti-C cross reactivity, including uncommon RHCE alleles that encode serine at residue p.103 (RHCE*ce.06.05, RHCE*ce.36), RHCE*ceTI, as well as those with RHD*DIIIa-CE(4-7)-D or RHD*01EL44, which is reported to express an altered C antigen [Smith A Transfusion 2023 63(S5):195A]. One donor for which DQS predicted C+ and serology found C-, carries a novel presumably null RHCE*Ce allele with c.922G >T (p.G310X). In three cases that type C+ on OV, medium-resolution molecular testing confirmed C- prediction of DQS; more work is required to rule out presence of an RHD-RHCE hybrid allele. In three cases DQS predicted C+ while serology typed the donor C- and three cases where DQS predicted C- while serology typed the donor C+. These six discrepancies were due to failure of DQS to accurately detect the RHCE 109bp insertion; two of these cases were in donors carrying RHCE*CE alleles.

Conclusions:

The rate of C antigen typing discrepancies between automated serologic and molecular platforms was very low (0.074%). Half of the discrepancies were due to RH variants, with 4% expressing weak antigens missed by serologic testing and 46% expressing variant antigens that cross reacted with anti-C.  The inability of DQS to accurately call the presence or absence of the RHCE 109 bp insertion that is typically associated with alleles expressing the C antigen was involved in 6 (22%) of the discrepancies. This is a known limitation of DQS [Young GJ Transfusion 2015; 55(S3):142]. It is unclear if the RHCE*CE allele(s) have different intronic sequences that might impact the ability of a molecular platform to amplify and detect the 109 bp insertion that is used to predict the C antigen. Testing of AA donors with both serologic and molecular platforms facilitated identification and resolution of a small number of C antigen typing discrepancies.